bmc international health and human rights · 2017. 8. 25. · page 1 of 12 (page number not for...

of 12(page number not for citation purposes)

BMC International Health and Human Rights

ResearchGender inequity and age-appropriate immunization coverage inIndia from 1992 to 2006 Daniel J Corsi1,2*, Diego G Bassani1, Rajesh Kumar3, Shally Awasthi4,Raju Jotkar1, Navkiran Kaur3, Prabhat Jha1

Address: 1Centre for Global Health Research, St. Michael’s Hospital, University of Toronto, Toronto, ON, M5C 1N8, Canada, 2Population HealthResearch Institute, McMaster University, Hamilton, ON, L8L 2X2, Canada, 3School of Public Health, Postgraduate Institute of Medical Educationand Research, Chandigarh, India, 4Department of Pediatrics, King Georges Medical University, Lucknow, Uttar Pradesh, India

Email: Daniel J Corsi* - [email protected]; Diego G Bassani - [email protected]; Rajesh Kumar - [email protected];Shally Awasthi - [email protected]; Raju Jotkar - [email protected]; Navkiran Kaur - [email protected]; Prabhat Jha [email protected]

* Corresponding author

from The fallacy of coverage: uncovering disparities to improve immunization rates through evidence.The Canadian International Immunization InitiativePhase 2 (CIII2) Operational Research Grants

Published: 14 October 2009

BMC International Health and Human Rights 2009, 9(Suppl 1):S3 doi:10.1186/1472-698X-9-S1-S3

This article is available from: http://www.biomedcentral.com/1472-698X/9/S1/S3

© 2009 Corsi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background: A variety of studies have considered the affects of India’s son preference on genderdifferences in child mortality, sex ratio at birth, and access to health services. Less research hasfocused on the affects of son preference on gender inequities in immunization coverage and howthis may have varied with time, and across regions and with sibling compositions. We present asystematic examination of trends in immunization coverage in India, with a focus on inequities incoverage by gender, birth order, year of birth, and state.

Methods: We analyzed data from three consecutive rounds of the Indian National Family HealthSurvey undertaken between 1992 and 2006. All children below five years of age with completeimmunization histories were included in the analysis.Age-appropriate immunization coverage wasdetermined for the following antigens: bacille Calmette-Guérin (BCG), oral polio (OPV), diphtheria,pertussis (whooping cough) and tetanus (DPT), and measles.

Results: Immunization coverage in India has increased since the early 1990s, but complete, age-appropriate coverage is still under 50% nationally. Girls were found to have significantly lowerimmunization coverage (p<0.001) than boys for BCG, DPT, and measles across all three surveys. Bycontrast, improved coverage of OPV suggests a narrowing of the gender differences in recent years.Girls with a surviving older sister were less likely to be immunized compared to boys, and a largeproportion of all children were found to be immunized considerably later than recommended.

Conclusions: Gender inequities in immunization coverage are prevalent in India. The lowimmunization coverage, the late immunization trends and the gender differences in coverageidentified in our study suggest that risks of child mortality, especially for girls at higher birth orders,need to be addressed both socially and programmatically.

Abstract in Hindi: See the full article online for a translation of this abstract in Hindi.

Open Access

Abstract in HindiSee Additional file 1 for a translation of the abstract tothis article in Hindi.

Background In India there is a well-documented history of son prefer-ence [1-4]. A growing body of literature has examined theimpacts of India’s son preference on child survival, juvenilesex ratio, and the numbers of ‘missing’ women [5-10]. Thereis evidence that the son preference in India and other SouthAsian countries contributes to disadvantage in women’shealth throughout the life course [11]. Disadvantage for girlsin India begins with a reduced chance of being born at all,and those who are born face lower access to preventive careand treatment of disease compared to boys [6,11,12].

Girls born in India have a 40% greater risk of ill healthcompared to boys and are less likely to access health care,in particular immunization [11,13,14]. Boys, however, aremore likely than girls to die in the first month of life fromperinatal conditions, such as birth asphyxia and birthtrauma. Only two other conditions (unintentionalinjuries and congenital anomalies) are more commonamong boys than girls. Beyond these causes, and contraryto the trends observed in most of the world, in India moregirls than boys die of acute respiratory diseases, infectiousand parasitic diseases, and viral infections [15,16].

Many of the deaths among India’s children are avoidable,including those from the childhood cluster of vaccine pre-ventable diseases (especially measles and tetanus), malaria,diarrhoea caused by organisms such as rotavirus, and acuterespiratory infections caused by Streptococcus pneumoniaand Hæmophilus influenza type b (HiB). Recent data showsthat immunization – long established worldwide as a high-ly cost-effective lifesaver – still reaches only a minority ofIndia’s children, even after the substantial improvements invaccination coverage against measles and polio. To makematters worse, girls are especially vulnerable due toinequities in access to immunization coverage [17].

Preference for sons in India has been noted to vary acrossregions, levels of fertility, and order of birth [1,18-20]. Awide variety of studies have examined the affects ofIndia’s son preference on child mortality and India’s sexratio in light of changing fertility patterns and concern for‘missing’ women. Less research has focused on the influ-ence of India’s son preference on gender inequities inaccess to health care, specifically immunization, and howthis may have varied with time and across regions. In thisarticle we investigate the presence of gender inequities interms of access to timely immunization coverage. We willfocus on trends in gender inequities at the national level,by birth order, and by state of residence using data col-lected from 1992 to 2006.

Previous researchInequities in immunization coverage by gender have beenshown to exist throughout India [12]. Of the 17 majorstates, 10 have demonstrated inequity in full immuniza-tion coverage against girls. Even states that perform wellin immunization coverage struggle with considerably dif-ferent immunization rates between boys and girls [20]. Asearch of available literature yielded several studiesreporting lower immunization coverage among girls ascompared to boys. A study of more than 4000 ruralIndian children in 1993-1994 indicated that fewer than55% of children were fully vaccinated and that girls had a5% lower coverage compared to boys [17]. In 1992, Bonuet al evaluated vaccination coverage among children aged12-35 months before and after a three-year governmentvaccination-awareness program in rural areas of fournorth Indian states. Prior to the program, girls were foundto be at a disadvantage compared to boys and the differ-ences in coverage by gender persisted following theprogram’s completion [21]. Other studies reviewed indi-cated lower immunization coverage for girls compared toboys, although differences were non-significant [22,23].

We compiled data from these four studies comprisingnine sub-samples (based on a combination of differentage groups and antigens) to obtain an overall ratio of cov-erage (girls vs. boys). We observed an overall coverageratio estimate of 0.93 (95% CI: 0.90, 0.9, Figure 1) – indi-cating that among these studies girls were 7% less likely tobe immunized when compared to boys (p<0.001).

MethodsData sourcesThis study uses data from three consecutive rounds of theIndian National Family Health Survey (NFHS) [24-26].The International Institute for Population Sciences coor-dinated each round of the survey with support fromseveral international organizations. The three cross-sec-tional surveys were conducted during 1992-93, 1998-99and 2005-06. A summary of the coverage and target pop-ulation for each round is presented in Table 1. Thesampling, questionnaire structure, and content of theNFHS surveys follow what has been adopted by theDemographic Health Surveys (DHS) in other developingcountries. The NFHS uses nationally representative area-based sampling frames in each survey [27]. The NFHSproduced high response rates in all states. Details of thesurvey methodology and response rates have been pub-lished for each round of the survey [24-26].

Sample for analysisOur sample for analysis includes all children below fiveyears of age with complete immunization histories (N =121,100). The 1998-1999 survey only included childrenup to 35 months of age at the time of the survey. About

BMC International Health and Human Rights 2009, 9(Suppl 1):S3 http://www.biomedcentral.com/1472-698X/9/S1/S3


3% of children were excluded due to missing data onimmunization coverage. Total sample sizes of childrenunder five along with analysis samples for each round ofthe NFHS are detailed in Table 1.

Indicators and measuresWe defined immunization coverage as up to date, age-appropriate immunization coverage. Standard indicatorsof immunization coverage are based on the percentage ofchildren who have accumulated the required number ofvaccines by a certain age, regardless of timeliness. Age-appropriate vaccination has been shown to be animportant component of infection control [28,29] by

reducing transmissibility in susceptible populations[30,31] and by increasing the probability of survival [32].We determined age-appropriate immunization coveragefor each antigen using a combination of data from thechild’s immunization card and maternal recall whencards were unavailable. Previous studies have demonstrat-ed that maternal recall can be a robust estimation ofimmunization coverage in settings where completerecords are not available [33].

Immunization information was available for the follow-ing antigens: bacille Calmette-Guérin (BCG), oral poliovaccine (OPV), diphtheria, pertussis (whooping cough)and tetanus (DPT) vaccine, and measles vaccine. We con-sidered children age-appropriately immunized if they hadreceived all immunizations for their age according to theWHO’s Expanded Program on Immunization (EPI)immunization schedule. Modelled on the WHO guide-lines, the government of India’s Universal ImmunizationProgram (UIP) was introduced in 1985 and includes onedose of BCG (at birth), three doses of OPV and DPT (at 6,10, and 14 weeks), and one dose of measles (at ninemonths) [34]. India’s EPI/UIP schedule used for our age-appropriate classification is detailed in Figure 2.

Using the child’s age in months and the EPI schedule, acomposite binary variable indicating EPI complete (yes orno) was created to represent the overall age-appropriateimmunization status of each child as follows:

0-1 month: child was considered age-appropriatelyimmunized (EPI=1) if they had received BCG;



Table 1 - Overview of India’s National Family Health Survey (NFHS).

Survey Phase

1992-1993 1998-1999 2005-2006 (NFHS-1) (NFHS-2) (NFHS-3)

Sample size (women) 89,777 91,000 230,000Age group (women) 13-49 15-49 15-49Data collection period

Start April 1992 Nov 1998 Dec 2005Finish Sept 1993 Dec 1999 Aug 2006

Number of states 25 26 29Living cohorts 1988-1993 1996-1999 2001-2006Total sample (living 45,275 30,821 48,468

children under 5)Sample included in analysis 43,732 29,669 47,709Reference period for immunization coverage

Start Jan 1988 Jan 1996 Jan 2001Finish Aug 1993 Dec 1999 Aug 2006

Figure 1 - Girl-to-boy ratios of immunization coverage and combined estimates derived from previouslypublished studies.The girl-to-boy immunization coverage ratios are based on the results of studies where data was available for calculation ofpooled estimates. A CR of less than 1.0 indicates higher coverage in favour of boys.

2-3 months: child was considered age-appropriatelyimmunized (EPI=1) if they had received BCG and twodoses of OPV and DPT (one dose of OPV and DPT if agedtwo months);

4-8 months: child was considered age-appropriatelyimmunized (EPI=1) if they had received BCG and threedoses of OPV and DPT;

9 months and older: child was considered age-appropri-ately immunized (EPI=1) if they had received BCG, threedoses of OPV and DPT, and one dose of measles.

To study the influence of birth order and gender of theolder siblings, we calculated the birth order and gender ofeach child in relation to the birth order and gender oftheir siblings. In a subset of children for whom the com-plete date of birth (day, month, and year) was known andcomplete date of immunization was recorded on theimmunization card, we calculated age of vaccine receiptfor all EPI antigens.

Data analysisThe analyses in this paper are primarily descriptive andpresent gender differences in immunization coverage byantigen, birth order, year of birth, and state across thethree rounds of the NFHS survey. Using the composite EPI

age-appropriate variable, we examined gender differencesin coverage by birth order and sibling composition. Wealso examined the change in age-appropriate coverage bybirth cohort (based on year of birth) for each EPI antigen.Births that occurred near to the time of the survey areexcluded from the cohort analyses in order to preventunderestimates of coverage due to reduced opportunity toreceive complete immunizations among these children.We instead report proportion of children not immunizedfor these birth cohorts. Gender differences in immuniza-tion coverage are presented at the state and national level.Sampling weights were used for all analyses. We tested dif-ferences between proportions using t statistics. Data weremanaged and analyzed using Stata (version 10) statisticalsoftware [35].

Results Gender inequities in age-appropriate immunizationcoverageImmunization coverage has increased in India since1992-1993 (Figure 3), but age-appropriate EPI coverageremains below 50% nationally for both boys and girls.Coverage of OPV has improved substantially according tothe 2005-2006 data, but progress has not been as markedfor DPT and Measles.

At the national level, age-appropriate BCG coverageamong boys increased from 58.2% (95%CI 57.4; 59.0) in1992-1993 to 76.0% (95%CI 75.3; 76.7) in 2005-2006.BCG coverage for girls increased from 54.9% (95%CI54.1; 55.7) in 1992-1993 to 73.1% (95%CI 72.3; 73.8) in2005-2006. Differences in BCG coverage between girlsand boys indicate that girls still have lower access to BCGat the national level (p<0.001, in all three periods).

The percent of boys with age-appropriate OPV coverageincreased from 50.1% (95%CI 49.3; 50.8) in 1992-1993to 78.7% (95%CI 78.1; 79.4) in 2005-2006 and from47.0% (95%CI 46.2; 47.8) in 1992-1993 to 77.5%(95%CI 76.8; 78.2) in 2005-2006 for girls. The most dra-matic increase in age-appropriate OPV coverage hashappened between 1998-99 and 2005-2006, but compar-ison of the coverage for boys and girls at the national levelstill shows significantly lower OPV coverage (p=0.012)among girls.

Coverage of DPT has shown less improvement. Accordingto the 1992-1993 data, age-appropriate DPT coverageamong boys was 48.5% (95%CI 47.7; 49.3), reaching52.4% (95%CI 51.4; 53.3) in 1998-1999 and climbingonly a few percent points in 2005-2006 to 55.7% (95%CI54.9; 56.5). Coverage for girls has increased from 45.1%(95%CI 44.3; 46.0) in 1992-1993 to 49.8% (95%CI 48.8;50.7) and to 53.3% (95%CI 52.5; 54.2) in 1998-1999 and2005-2006, respectively. Girls, however, are experiencing



Figure 2 - Recommended expanded program ofimmunization (EPI) schedule for India.*OPV-0 is an additional dose of polio given at birth, but isnot part of India’s national immunization program. †Dashedlines indicate ages (1,2,3,5, and 9 months) used indetermining appropriate immunization coverage within thechild’s first year. Children were considered to have age-appropriate EPI coverage if they had received all antigensrecommended for their age.

significantly lower coverage rates compared to boys(p<0.001, in all three periods).

Measles, an antigen that requires over 95% coverage tostop transmission in a population, still has very low cov-erage rates in India. For boys, there has been an increasein age-appropriate coverage from 43.9% (95%CI 43.0;44.8) in 1992-1993 to 58.6% (95%CI 57.8; 59.5) in2005-2006. Age-appropriate coverage among girlsincreased from 40.7% (95%CI 39.9; 41.6) in 1992-1993to 56.1% (95%CI 55.2; 57.0) in 2005-2006. Differencesin coverage by gender are still significant (p<0.05) withgirls experiencing lower coverage than boys. Large increas-es can still be made in the overall levels of coverage ofmeasles vaccination.

As can be noted from the gender specific coverage ratesreported above, gender differences in vaccination cover-age are consistent and significant, but do not appear tohave increased over time. Girls’ coverage for every antigenlags behind boys’ coverage in all years, but neither gender

displays acceptable age-appropriate coverage levels forany antigen (Figure 3).

Inequities in age-appropriate immunization coverage dueto birth orderBirth order and family composition is an important pre-dictor of vaccination coverage. Higher birth order isassociated with a lower probability of being age-appropriately immunized. Despite the increase in rates ofage-appropriate immunization coverage over time, thegender gap has not been reduced. Girls are much lesslikely to be up to date with their immunization at anygiven age. Furthermore, the gap increases with birth orderunless a girl has an older brother. Girls who are born thirdto a family with two other girls are in the extreme ofimmunization disadvantage, when compared to boyswho are born third to families with two older girls. Only36.1 % (95%CI 33.3; 39.0) of third born girls with twoolder sisters are appropriately immunized for their agecompared to 45.0% (95%CI 42.3; 47.8, Figure 4) of thethird order boys with two older sisters. Higher birth order



Figure 3 - Percent of children with age-appropriate coverage, by gender, antigen, and date of survey.Period of NFHS Surveys: 92-93 (NFHS 1), 98-99 (NFHS 2), and 05-06 (NFHS 3).Vertical bars represent 95% confidenceintervals (CI).

children (third or higher) with mixed gender sibling (i.e.brothers and sisters) composition have very low age-appropriate immunization coverage, not reaching 30%.Still, girls are experiencing lower coverage rates comparedto boys.

This trend is inverted for third order births following twoolder boys (Figure 4). In this situation, the gender gap isnot as pronounced. Like all children at higher birthorders, these children are vulnerable to lower age-appro-priate immunization coverage. Indeed, the coverage ratesfor these children are lower in 2005-2006 than ratesobserved for first order children born more than 10 yearsbefore (i.e. in 1992-1993, Figure 4).

Trends in gender inequities over timeWe present trends in age-appropriate immunization cov-erage by birth cohort and antigen in Figure 5.Age-appropriate immunization by gender is presented forchildren born during the years between 1988 and 1993,between 1996 and 1999, and between 2001 and 2006inclusive. For children born during the survey years(1992-93 for NFHS-1, 1997-98 for NFHS-2, and 2005-6for NFHS-3), we present the proportion of children notage-appropriately immunized in Figure 6.

Age-appropriate coverage of BCG increased over thecohorts analysed for both boys and girls. In the majorityof cohorts, girls received lower age-appropriate BCG cov-erage than boys. This difference remained significant forchildren born in 2003 and 2004 (p<0.05), Figure 5. Forchildren where the date of immunization was recorded,the median age of receiving BCG decreased from 11.3weeks in 1990 to 7.0 weeks in 1996 and to 4.7 weeks in2003. For boys the decrease has been from an age of 11.6weeks in 1990 to 6.7 weeks in 1996 and to 4.6 weeks in2003. The WHO/EPI recommendation is that BCG beadministered at birth or first contact with health services.Our results indicate a considerable delay in BCG coveragefor both boys and girls (see web appendix in Additionalfile 2). The proportion of boys and girls born during thesurvey years who were not vaccinated are presented inFigure 6. These estimates are cross-sectional and shouldbe interpreted taking into account future opportunity forimmunization.

Age-appropriate coverage of OPV has improved from1988 to 2004 and similar coverage for both boys and girlshas been achieved in recent birth cohorts. Where infor-mation is available, it is noted that the age of receivingOPV-1 has been reduced from a median of approximate-ly 13 weeks in 1990 (boys and girls) to less than nineweeks in 2003, near to the recommended age of six weeks.The median time to the second and third polio doseremains above the WHO recommendation of 10 and 14

weeks, respectively (Additional file 2). There has been areduction in median age for receiving OPV-2 from 18weeks (for all children) in 1990, to 15 weeks in 2003. Themedian age to receive OPV-3 was 24 weeks for boys and33 weeks for girls in 1990. This has been reduced to 21weeks for boys and 27 weeks for girls in 2003. Girls bornin 2003 received OPV-3 more than 13 weeks later thanrecommended.

Important gender inequities in DPT coverage were notedfor children born in 1989, 1991, and 1996. Theseinequities have not persisted for children born in 2001-4(Figure 5). Overall levels of DPT coverage have, however,remained below 60% for boys and girls. The median agefor receiving the first DPT dose was seven weeks later thanrecommended for girls and boys in 1990. This hasreduced to three weeks later than recommended for chil-dren born in 2003 (Additional file 2). Boys and girls bornin recent years are also receiving the second and third DPTdoses earlier (Additional file 2).

There has been an increase in age-appropriate measlesimmunization coverage for boys and girls born in 2004compared with children born in 1988. A difference incoverage between girls and boys is still apparent, even forchildren born in recent years. Girls born in 2004 had sig-nificantly (p<0.05) lower coverage than boys born in thesame year. Both girls and boys continue to receive themeasles vaccine about six weeks later than recommended(Additional file 2).

Age at immunizationAcross all antigens, the distributions of age at immuniza-tion demonstrate significant positive skew (Additionalfile 2). The children at upper tails of these distributionsare being immunized extremely late. This indicates that asizable amount of children are behind the recommendedimmunization schedule for all of the evaluated antigens.The median age at immunization appears to be decreas-ing, suggesting an increase of children being vaccinatedon time for those born in recent years. Considerablevariability exists in the sample sizes available in each birthcohort due to incomplete data on child’s date of birthand/or dates of immunization. There does not appear tobe large gender inequities in terms of age at immuniza-tion, but a sizeable number of children are behind therecommended immunization schedule for all the evaluat-ed antigens (Additional file 2).

Gender inequities in immunization coverage by state andregion in IndiaSubstantial variation in levels of EPI coverage for boys andgirls exists between states in India (Figure 7).Immunization coverage was not strongly correlated withgender inequities in coverage (Pearson’s r = 0.06). Six out





Figure 4 - Percent of children with age-appropriate coverage of EPI vaccines, by birth order, gender ofprevious birth(s), and survey period.Overall (EPI) age-appropriate immunization coverage for girls and boys at each birth order (column 1) and sibling gendercomposition of previous birth(s) (column 2) is represented by the bar chart for each year of the survey period (column 3).†Percentages in brackets (column 1) are the proportion of all births represented by each category of birth order and siblinggender in the 2005-6 survey. Horizontal bars represent 95% CI.

of the fifteen states that perform above the national aver-age in immunization coverage are also among the eightstates with the lowest ratios of girl-to-boy immunizationcoverage (Figure 7). Nationally, the girl-to-boy ratio ofimmunization coverage is 0.95 (p<0.001), demonstratingthat 5% fewer girls than boys are fully immunized.

The three states with the lowest girl-to-boy ratios of immu-nization coverage are the northern states of Punjab,Haryana, and Bihar. In these states, between 10 and 14%fewer girls than boys are fully immunized. Other northernstates, including Uttar Pradesh and Delhi, demonstrate lowgirl-to-boy coverage ratios (0.92, and 0.93, respectively)



Figure 5 - Trends in age-appropriate immunization coverage of EPI antigens in India by antigen, gender, andyear of birth.Years of birth covered by NFHS surveys are indicated by solid points and lines. Dashed lines indicate periods where data isnot available.Vertical bars represent 95% CI.

although these differences in coverage are not statisticallysignificant. The southern Indian states of Tamiml Naduand Keral demonstrate high levels of immunization cov-erage (80.5% and 75.4%, respectively) and near equity incoverage by gender. Gender inequities in immunizationcoverage are not clearly determined by a north southdivide. Two northern states, Himachal Pradesh andRajasthan, demonstrate an immunization coverage ratiothat favours girls (Figure 7). Additionally, the southernstate of Andhra Pradesh demonstrates a large difference incoverage favouring boys (0.93, p=n.s.). The northeasternstates of Jharkhand and Assam demonstrate the largestcoverage ratios in favour of girls, but are among the stateswith the lowest coverage rates.

Using proportions of children not appropriately immu-nized for their age in 2005-6 and population data fromthe 2001 Census [36], we estimate that, nationally, up to31M girls and 32M boys aged 0 to four are lacking age-

appropriate immunization coverage in India. We presentestimated numbers of children lacking age-appropriatecoverage for boys and girls aged 0 to four in 21 states inFigure 7. Nationally, and for many states, the absolutenumber of boys exceeds the number of girls lacking com-plete age-appropriate immunization coverage. This is dueto gender imbalances in India’s population structure,leading to fewer girls than boys at these ages [6,36].

Discussion Using three nationally representative surveys from India,our findings indicate that, at the national level, girls havelower immunization coverage than boys. We noted thatgirls at higher birth orders and with older sisters are atgreater risk of missing antigens compared to boys of thesame birth order and sibling gender composition.Analyses of children by birth year revealed consistentlylower coverage of BCG and measles for girls compared toboys, while also suggesting that gender inequities in OPV



Figure 6 - Percentage of children without age-appropriate immunization coverage in India at the time of eachNFHS survey, by antigen and gender.We calculated the percentage of children born during each survey period who were not appropriately covered for eachantigen according to their age.Vertical bars represent 95% CI.

and DPT coverage may have narrowed for children bornin recent years. Gender inequity in immunization cover-age was found to be greatest in three northern Indianstates, consistent with previous findings of strong sonpreference in these areas [1,7,19]. The gender inequitiesin access to preventive care in India noted here are likelyto reflect, at least in part, the societal preference for sonsin India [11,20].

Our estimates are lower than those published in theNFHS reports as these reports do not account for the time-liness of vaccine delivery [24-26]. Analyses of age atimmunization indicated minimal differences betweengirls and boys for all antigens, suggesting a child's gendermay affect the decision to immunize, but not timelinessof coverage.

The use of age-appropriate immunization coverage ratesas a health indicator can aid in determining the conse-quences of not receiving timely vaccination. In ouranalysis, a substantial number of boys and girls werereceiving BCG between one and two years after birth.

Studies in the US indicate that the timeliness of immu-nization coverage can have a considerable impact onchild survival [37,38]. In Bangladesh, a study revealedtimely BCG immunization could reduce the mortality riskup to 40% for children vaccinated between 60 and 180days of life and up to 80% for children vaccinated withinthe first 60 days of life [32]. Similar results have beenshown for the effect of age-appropriate DPT immuniza-tion and the risk of pertussis [39]. Socioeconomiccharacteristics associated with delayed immunizationhave been rarely studied in India [40] and further researchis needed in this area.

Our finding that gender inequities in immunization cov-erage are not correlated with overall immunizationcoverage is consistent with previous studies [12,20].Several states with low girl-to-boy coverage ratios (e.g.Punjab, Haryana, and Delhi) are performing above thenational average in terms of overall coverage. Data fromthe DHS in India and other South Asian countries indi-cate that the gender differences in immunization coverageare observed among all socio-economic levels, with some



Figure 7 – Percentage of children with age-appropriate EPI coverage, girl-to-boy coverage ratio, and numberof children without age-appropriate immunization coverage in India (2005-2006), by state and gender.NFHS 3 data and state-level sampling weights were used for this figure. †Absolue number of children without age-appropriateEPI coverage, estimated from the 2001 Indian Census [34]. *CR, coverage ratio, girls vs. boys **p ≤0.001.

indication that differences may be larger among the richthan among the poor [12].

Gender inequity in access to health services is believed tobe one consequence of larger societal circumstancesacross South Asia that favour boys and leads to the mar-ginalization of women from a young age [11]. Anexploration of socioeconomic and sociocultural determi-nants of gender inequities in immunization coverage isbeyond the scope of this work. Readers are referred toother studies investigating the determinants of genderinequities in sex ratios [6,7], child mortality [5,41,42] fer-tility patters [1,2], and nutrition [17,20]. Furtherinvestigation into the determinants of continuing genderinequities in India’s immunization coverage is required.

Along with improved nutrition, immunization is a majortool for saving children’s lives, standing out, even in set-tings of very low income, as a highly cost-effective andefficient intervention [43]. Currently, in India, two-thirdsof the children who die of measles and other preventablechildhood diseases would have survived if they had hadaccess to immunization. The additional annual per capitacost necessary to reach 90% of Indian children with thesix basic vaccines already included in the national immu-nization program – diphtheria, tetanus, pertussis, TB(BCG), polio, and measles – would be less than threerupees (eight US cents) in the poorest states and even lessin wealthier states [44].

Our study shows that current inadequate levels of immu-nization coverage in India are only part of the problem.Gender inequity in access to health programs is responsi-ble for a considerable number of avoidable deaths. India’sstates could save even more lives by addressing deep-seat-ed social and cultural issues responsible for genderdiscrimination at the household level, where girls are seenas a burden and boys as a resource [41]. Campaigns toraise awareness of gender inequities in conjunction withimprovement in vaccine delivery strategies with a focuson timeliness of coverage may be a means to that end.Continuing research is needed in order to identify effec-tive social policies of reducing gender inequities in accessto immunizations and other health practices across Indiaand other parts of South Asia.

Conclusions In India, inequities in girls’ access to health services havepersisted in recent years. Our research has examined thisdisparity through a study of gender inequities in childimmunization coverage. Young girls are especially vulner-able to these gaps in coverage, but efforts need to be madeto increase overall immunization coverage for both girlsand boys. Over half of children under five are not fullyimmunized, and a large proportion of those immunized

are being immunized too late. These two factors can leadto increased transmissibility of infections, reducingprotective effects of immunization and contributing toavoidable child deaths. In 2005, over 60 million childrenwere not appropriately immunized in India.

List of abbreviations usedBCG – Bacille Calmette-Guérin; DHS – DemographicHealth Surveys; DPT – Diphteria, Pertussis and Tetanus;EPI – Expanded Program on Immunization; Hib –Hæmophilus influenza type b; NFHS – Indian NationalFamily Health Survey; OPV – Oral Polio Vaccine; UIP –Universal Immunization Program.

Competing interestsThe authors declare that they have no competing interests.

Authors’ contributionsDC and DB conducted the literature review, data analysisand drafted the manuscript. PJ, RK, SA, RJ, NK, DB, andDC designed and coordinated the study. All authors par-ticipated in the interpretation of the results and reviewingthe final manuscript.

Additional materialAdditional file 1 Abstract in Hindi. Available from:http://www.biomedcentral.com/content/supplementary/1472-698X-9-S1-S3-S1.doc

Additional file 2 Web Appendix: Distributions of time to immunization, by antigen,gender and year of birth.This web appendix presents the distributions (box andwhisker plots) of the time to vaccination for each antigenby year of birth. Boys and girls are presented separately.Dashed red lines indicate time of age-appropriate vaccineadministration. Red diamonds indicate the mean time tovaccination. The full data table is also presented. Childrenborn in 1994, 1995, 2000 and 2006 (for measles only),are not covered by the NFHS surveys and no data is pre-sented for these years. Available from:http://www.biomedcentral.com/content/supplementary/1472-698X-9-S1-S3-S1.doc

AcknowledgementsThis work was carried out with the aid of a grant from the InternationalDevelopment Research Centre (IDRC), Ottawa, Canada, as part of theCanadian International Immunization Initiative Phase 2 (CIII2).This initiativeis a project of the Global Health Research Initiative (GHRI). We thankLukasz Aleksandrowicz for his assistance in the literature review andpreparation of the manuscript.We also thank MEASURE DHS for makingthe NFHS India datasets available.



This article is published as part of BMC International Health and HumanRights Volume 9 Supplement 1, 2009: The fallacy of coverage: uncoveringdisparities to improve immunization rates through evidence.The CanadianInternational Immunization Initiative Phase 2 (CIII2) Operational ResearchGrants. The full contents of the supplement are available online athttp://www.biomedcentral.com/1472-698X/9?issue=S1.

References1. Bhat PNM, Zavier AJ: Fertility decline and gender bias in

northern India. Demography 2003, 40:637-657.2. Mutharayappa R, Choe MK, Arnold F, Roy TK: Son preference and

its effect on fertility in India. National Family Health Survey SubjectReports No. 3. Mumbai: International Institute for Populations Sciences;1997.

3. Cleland J,Verrall J,Vaessen M: Preferences for the sex of childrenand their influence on reproductive behaviour. Word FertilitySurvey, Comparative Studies, No. 27. Voorburg, Netherlands:International Statistical Institute; 1983.

4. Williamson NE: Sons or daughters: A cross-cultural survey of parentalpreferences. Beverly Hills, California: Sage Publications; 1976.

5. Arnold F, Choe MK, Roy TK: Son preference, the family buildingprocess and child mortality in India. Popul Stud 1998, 52:302-315.

6. Jha P, Kumar R,Vasa P, Dhingra N,Thiruchelvam D, Moineddin R: Lowfemale[corrected]-to-male [corrected] sex ratio of childrenborn in India: national survey of 1.1 million households. Lancet2006, 367:211-218.

7. Bhat PNM: On the trail of ‘missing’ Indian females. (Published intwo parts) Economic and Political Weekly 2002, 37:5105-5118, 5244-5263.

8. Guillot M: The dynamics of the population sex ratio in India,1971-96. Popul Stud 2002, 56:51-63.

9. Das Gupta M, Bhat PNM: Fertility decline and increasedmanifestation of sex bias in India. Popul Stud 1997, 51:307-315.

10. Sen A: Missing women: revisited. BMJ 2003, 327:1297-1298.11. Fikree FF, Pasha O: Role of gender in health disparity: the South

Asian context. BMJ 2004, 328:823-826.12. Pande RP, Yazbeck AS: What’s in a country average? Wealth,

gender, and regional inequalities in immunization in India.Soc Sci Med 2003, 57:2075-2088.

13. Thind A: Health service use by children in rural Bihar. J TropPediatr 2004, 50:137-142.

14. Filmer D, King EM, Pritchett L: Gender disparity in South Asia:comparisons between and within countries. Policy ResearchWorking Paper No. 1867.Washington, DC:World Bank; 1998.

15. Registrar General of India: Survey of causes of death (rural): annualreport 1998. New Delhi: Registrar General; 2000.

16. World Health Organization: The world health report 2006, Workingtogether for health. Geneva:World Health Organization; 2006.

17. Borooah VK: Gender bias among children in India in their dietand immunisation against disease. Soc Sci Med 2004, 58:1719-1731.

18. Bhat PNM, Zavier F: Findings of National Family Health Survey:Regional Analysis. Economic and Political Weekly 1999, 34:3008-32.

19. Dyson T, Moore M: On kinship structure, female autonomy, anddemographic behaviour in India. Population Devel Rev 1983, 9:35-60.

20. Pande RP: Selective gender differences in childhood nutritionand immunization in rural India: the role of siblings.Demography 2003, 40:395-418.

21. Bonu S, Rani M, Baker TD: The impact of the national polioimmunization campaign on levels and equity inimmunization coverage: evidence from rural North India. SocSci Med 2003, 57:1807-1819.

22. Elliott C, Farmer K: Immunization status of children under 7years in the Vikas Nagar area, North India. Child Care Health Dev2006, 32:415-421.

23. Moulton LH, Rahmathullah L, Halsey NA,Thulasiraj RD, Katz J,TielschJM: Evaluation of non-specific effects of infant immunizationson early infant mortality in a southern Indian population.Trop Med Int Health 2005, 10:947-955.

24. International Institute for Population Sciences: National Family HealthSurvey (MCH and Family Planning) India 1992-93. Bombay; 1995.

25. International Institute for Population Sciences and ORC Macro:National Family Health Survey (NFHS-2) 1998-99, India. Mumbai; 2000.

26. International Institute for Population Sciences and MacroInternational: National Family Health Survey (NFHS-3) 2005-06, India.Volume I. Mumbai; 2007.

27. Macro International: Sampling Manual, DHS-III Basic Documentation No.6. Calverton, Maryland; 1996.

28. Atkinson WL, Hadler SC, Redd SB, Orenstein WA: Measlessurveillance-United States, 1991. MMWR CDC Surveill Summ1992, 41:1-12.

29. Siedler A, Hermann M, Schmitt HJ, Von Kries R: Consequences ofdelayed measles vaccination in Germany. Pediatr Infect Dis J2002, 21:826-830.

30. Gay NJ: The theory of measles elimination: implications forthe design of elimination strategies. J Infect Dis 2004, 189(Suppl1):S27-35.

31. Gidding HF,Wood J, MacIntyre CR, Kelly H, Lambert SB, Gilbert GL,McIntyre PB: Sustained measles elimination in Australia andpriorities for long term maintenance. Vaccine 2007, 25:3574-3580.

32. Breiman RF, Streatfield PK, Phelan M, Shifa N, Rashid M, Yunus M:Effect of infant immunisation on childhood mortality in ruralBangladesh: analysis of health and demographic surveillancedata. Lancet 2004, 364:2204-2211.

33. Langsten R, Hill K: The accuracy of mothers’ reports of childvaccination: evidence from rural Egypt. Soc Sci Med 1998,46:1205-1212.

34. Madhavi Y: Vaccine policy in India. PLoS Med 2005, 2:e127.35. StataCorp: Stata Special Edition, version 10.1. College Station, Texas;

2008.36. Registrar General of India: Provisional Population Totals, Census of India

2001. New Delhi: Registrar General; 2001.37. Luman ET, Barker LE, McCauley MM, Drews-Botsch C: Timeliness of

childhood immunizations: a state-specific analysis. Am J PublicHealth 2005, 95:1367-1374.

38. Luman ET, Barker LE, Shaw KM, McCauley MM, Buehler JW, PickeringLK: Timeliness of childhood vaccinations in the UnitedStates: days undervaccinated and number of vaccinesdelayed. JAMA 2005, 293:1204-1211.

39. Briand V, Bonmarin I, Levy-Bruhl D: Study of the risk factors forsevere childhood pertussis based on hospital surveillancedata. Vaccine 2007, 25:7224-7232.

40. Ughade SN, Zodpey SP, Deshpande SG, Jain D: Factors responsiblefor delayed immunization among children under 5 years ofage. J Indian Med Assoc 2000, 98:4-5,14.

41. Das Gupta M: Selective discrimination against female childrenin rural Punjab. Population Devel Rev 1987, 1:77-100.

42. Kishor S: Gender differentials in child mortality: A review ofthe evidence. In Women’s health in India: Risk and vulnerability. Editedby Das Gupta M, Chen LC, Krishnan TN. Bombay: Oxford UniversityPress; 1995:19-54.

43. Bloom D, Canning D,Weston M: The Value of Vaccination. WorldEconomics 2005, 6:24.

44. Jha P, Laxminarayan R: Choosing health: an entitlement for allIndians. Toronto: Centre for Global Health Research; 2009.



bmc international health and human rights · 2017. 8. 25. · page 1 of 12 (page number not for...

Documents